INDUCED RADIOACTIVITY IN ACCELERATORS M. Barbier CERN INTRODUCTION which can indicate the exact dose in rads. Also, the detector used should be fairly omnidirectional, A lot of attention has been paid recently as will come from all to the activation of accelerators. The radioactivity directions in a machine hall. induced by the flux of particles during the operation of the machine remains after the latter has stopped running. It constitutes a permanent danger to staff and restricts access to the equipment for operation, maintenance, repairs and alterations, in a way that is already being felt. It is beginning to be realized that it is the activation level that will set a practical limit to the beam intensities of future machines. Accelerators are intended for constantly changing experimental work, involving frequent modifications in parts of the machine itself. For a series of investigations the accelerator cannot be operated as a box delivering a beam of primary particles. Depending upon the type of experiments in view, one must expect to run into a «barrier of radioactivity» that will dictate the maximum intensity of the internal beam if the machine is to remain a useful nuclear physics tool. The principal methods for studying induced radioactivity will be outlined. These include measurements on a radioactive machine, experiments to establish the properties of various materials, and the calculation of radiation fields. Some experimental data collected at the CERN Synchro-Cyclotron will be given as an illustration.

1. ACCELERATOR ACTIVATION SURVEY

The first experiment to perform in order Fig. 1. Lead protected counter. to study induced radioactivity on an accelerator is to measure the radiation level and The next step is to find out which parts its decay with time at different points in the of the machine are most active. For this it is machine hall. necessary to place a collimator in front of the Any gamma radiation detector can be used: counter and shield it on all sides. Fig. 1 shows Geiger-Müller tubes, scintillators, films, ionization a lead protected counter, which can be pointed chambers. However, as their response in any desired direction. If this device is varies with energy, they should be calibrated equipped with a NaJ crystal activity spectra for each radiation spectrum or location against from specific parts of the machine can be recorded. the tissue equivalent ionization chamber,

1386 Fig. 2. Spherical plastic scintillator.

Another task which has to be done frequently In this connexion, it is a good idea is to map the intensity of gamma radiation also to measure the dependence of the activity in very hot and possibly inaccessible spots, for upon the depth of the activated material. instance between the pole pieces of a cyclotron. This can be found by examining screws or A plastic scintillator is recommended for very bolts which were perpendicular to the surface. high counting rates. The shape of the scintillator Routine surveys should include radiation can be spherical, as the counting rate field measurements in the principal locations is then independent of the distance from a plane each time maintenance is carried out during of uniform activity. A plastic sphere, with a shut-down, for instance, in front of thin a hollow light guide and a photomultiplier windows of the vacuum tank. In this way the mounted on the end of a long pole, as used general variation of the activity of the machine at CERN, is shown in Fig. 2. with time can be followed, and it can be stated For a better understanding of the distribution whether the activation level has reached of activity in the machine, calculations saturation or whether it will grow in the future, of the flux falling on the pole pieces or accelerating and levels to be expected for a given increase cavities should be made, taking into of the internal beam can be predicted. account target position, mean scattering angle, multiple traversal, and focusing action of II. THE PROPERTIES OF VARIOUS magnetic fields. Practical measurements MATERIALS WITH RESPECT TO ACTIVATION of fluxes in hot regions can be made by placing small samples or wires in the machine to be The high energy cross-sections activated. The distribution of activity along for building one element from another the wires shows the geometrical flux distributionsby the. impact of a particle have been

1387 studied to a certain extent by the radiochemists. Definite information can only be expected They form the basis of preliminary from samples which have been irradiated activation calculations, but need to be complementedto saturatio n point. In that case the relative by other experimental investigations position of the decay curves for the various on irradiated samples of various materials. materials and their shape could be completely When irradiating samples, several things different from those presented here. Long-term must be borne in mind. First, the irradiation irradiation experiments are now in progress time must be of the same order as the half life at CERN. of the newly formed elements which it is wished to study. Then the size of the sample III. THE CALCULATION must be chosen according to the scope of the OF THE ACTIVITY PER GRAM experiment. Both small and large samples AS A FUNCTION OF THE FLUX should be used. From small samples one can get the specific activity n in counts per gram The purpose of this work is obviously to make and sec in the solid angle 4π, the decay with it possible to calculate in advance the field time, the gamma energy spectral distribution, of radiation in a given installation which has and the beta-activity. The counter efficiency been exposed ot the action of a given flux should be determined with known sources of high-energy particles. Once the cross-sections of gamma radiation of several energies. Large are known, one can first determine the activity samples are also useful because they give per gram as a function of the flux received. an opportunity of also measuring the absorption Supposing the sample has been exposed of the gamma radiation by the active to a constant flux Φ during an irradiation time material itself. A cylindrical shape is suitable Ti, then the number of active atoms of the with a section equal to the section of the element ν per gram of matter at an instant detector and a length equal to two or three t will be equal to Ti t-τ times the absorption length of the main gamma N N = Φσ Tν dτ, rays emitted. The sample should be placed ν ν A on the axis of the detector, but some distance ∫0 e 2 away from it in order to minimize geometrical where σν cm is the cross-section for the effects due to the length of the sample. With formation of the element ν, Φ sec-1.cm-2 these samples one can measure the danger the incident flux of particles, N/Ath e number parameter n/x (x being the mass absorption of atoms per gram of the element exposed to the flux (N = 6×l023 is the Avogadro coefficient), the shape of the spectrum actually number and A the atomic weight in grams), received from large active bodies and the and Tν the time constant of decay to 1/e of radiological dose to which this spectrum corresponds.the activity formed (1.44 times the half-life). It is convenient to place these large The factor ΦσνN/A indicates the number of samples against the tank wall outside the atoms formed per unit of time and this number vacuum chamber and to measure them every decreases exponentially with the time constant maintenance day in order to record the growth Tν from the moment of formation Τ up to the of activity. present moment t, the time being counted from To illustrate all this, Figs. 3 and 4 show the beginning of the exposure. What interests decay curves of various materials exposed us is the activity itself, i. e. the number to protons (in the cyclotron tank, at some of atoms decaying per second. This can be obtained distance from the target but at the same radius) by deriving the above expression with and (outside the tank in a forward respect to time. The total of all the elements direction). The irradiation times were three formed should also be made on the ν indexes, months and six weeks respectively. The mean which gives: efficiency of the counter in the energy range was 8% so that the counts/gram sec still have dNν 23 1 to be multiplied by 12.5 to give the specific = 6 × 10 Φ × activity n in photons/gram sec in 4π. Although Σ dt A they are interesting, these curves do not represent ν the exact situation in accelerators, Τ Τi Σ σ e (1 - e where the irradiation time is much longer. × ν ν Tν Tν ),

1388 1389

Fig. 3. Decay of samples exposed to protons. Proton exposure Fig. 4. Decay of samples exposed to neutrons. exposure 3 months. 6 weeks. where T = t - T1 is the time which has angle subtended by the active element of elapsed since the end of the exposure. volume. Since the active masses may be spread over IV. THE CALCULATION OF THE FIELD a large area, it is advisable to assume that the OF RADIATION detector is spherical and consequently presents the same cross-section in all directions. This When one has a block of active matter simplifies the geometry and also the calculations. of a certain volume it is evident that part The number of photons reaching the of the radiation coming from the interior will sphere per second is then given by the volume be absorbed before reaching the surface. This integral absorption should be taken into account when 2 N = πa • e- • n dV, evaluating the radiation to which the staff are ∫4π x2 exposed. As a first approximation, it is sufficient ν to consider the mass absorption coefficient x (g-1.cm2) for narrow beams and where a is the radius of the detector sphere, to suppose an exponential decay of the intensity x the distance between the sphere and the active element of volume, x the portion of this NΓ of the radiation with the number of g/cm2 through which it has to pass: distance which is inside the active element and V the active volume covered by the integral. The activity n may be a function of position. -0x Nγ = N0γe V. TWO EXAMPLES OF FIELDS where x is the path and the density. OF RADIATION Once the activity per gram n and the absorption coefficient of the material irradiated are known, it is onlyOn a questione case whicof geometryh occur s in practice is that to calculate the field of radiation existing of a field of radiation in the vicinity of a flat outside the active mass. Each element of volume active layer of indefinite extent. One can give dV sends a quantity of photons ndV into the solid angle 4π. Let n represents the number of photons emitted in 4Π per gram per second. To determine it, the mode of decay must be taken into account. If only one gamma is emitted per decay, n will be equal to dNν dt for each element ν, and it there are two it will be doubled. If a is emitted which immediately recombines to give two 510 keV photons, it will also be doubled, etc. It is seen that n can be calculated from Σ dNν by taking dt into account the mode of decay for each fission product and multiplying the corresponding corss-section σν by an appropriate coefficient. It can also be measured experimentally on an active sample. It is, at T = 0 Fig. 5. Radiation field of an active and for infinitely long exposure layer. 1 n = Φ×6·1023× A Σσν. this layer a thickness T. Fig. 5 is a schematic ν diagram of the situation. Let D be the distance The photon flux issuing from ihe element between the detector and the layer. As the of volume evidently decays in inverse proportion layer covers an indefinite area the measurement to the square of the distance and in addition will not depend on D, but D must be is exponentially attenuated along that introduced for purpose of calculation. One part of its path which lies inside the radioactive may use spherical co-ordinates with an axis material. The flux received by the perpendicular to the layer, α being the site detector is obviously proportional to the solid complement.

1390 The element of volume becomes dV = is generally much greater than the length of =2πx2 sin α dαdx. One obtains absorption of gamma rays, the active area can D be taken to extend to infinity in the direction N 1 - (x = n cos α )dV = of the axis. One obtains πα2 ∫4π x2 e ν R ∞ 1 π/2 N 1 2 2 n πa2 = n 4π(r2+z2) e x √r +z 2πr dr dz, = (1 - e-T/cos α) sin α dα. ∫ ∫ 2 ∫ o o α where X = 1/. In the event of the thickness T being infinite After integration as a function of r one has (all the half space being active) one finds N n πα2 = 2 iN 1 n -1. -2 [1- 2 = sec cm . πα 2 ∞ 2 2 - Ei(- R +z z - 2 )], When the layer has a finite thickness, it can ∫ √ X ) d ( X be integrated in parts and one finds 0 T 1 an expression which should be graphically IN 'n X T T 2 = Ei(- integrated for different R/X parameter values. πα 2 [1 - e X X )], where X is the length for attenuation of the radiation to 1/e and where the function

∞ e-t - Ei(- x) = dt ∫ t X is well known in sine integral theory. Fig. 6 represents the expression is square brackets which shows how the field of radiation grows with the thickness of the active layer. Another case which occurs in practice is that of an absorber for a high-energy particle beam. Let πR2 be the cross-section of the beam presumed to be circular and let it fall perpendicularly Fig. 7. Radiation field versus beam radius.

Fig. 7 gives the shape of the quantity in sequare brackets as a function of R/X, namely the variation of the intensity of the radiation with the radius of the beam cross-section (the beam flux but not its total intensity is presumed to remain constant). It is seen that one reaches about the same value for n/2a s in the previous case, as soon as R/Xexceed s 1.

VI. THE n/ RATIO FOR DIFFERENT Fig. 6. Radiation field versus thickness fo MATERIALS active layer. It was seen in the previous paragraph that on to the absorber. The value which it is necessaryth e field of radiation is determined by the to calculate is the intensity of the activation ratio n/ in addition to the geometrical conditions. radiation on the surface of the absorber Knowing the flux Φ, it can be calculated at the centre of the area on which the beam for different elements by using the radiochemical falls. Cylindrical coordinates (r, z) centered data available and taking into account in the beam on the surface of the absorber will the mode of decay and the energy of the gammas be used. Since the penetration of the protons emitted. The absorption coefficient depends

1391 very little on the element converned in a large VII. CONCLUSION area but it depends a great deal on the energy Although this work is rather incomplete, of the photon. The energy of the photons emitted it nevertheless gives some idea of the phenomena by each fission product should therefore occurring when atomic installations are be considered and the cross-section σν divided activated. The results already obtained make by the appropriate value of . it possible to a certain extent to predict the This compilation was done for several elementfields s of radiation to be expected when the l and the quantity ∑ σν , which primary fluxes and the geometrical arrangement A ν ν of the activated elements are known. we call the «danger parameter for the material, It will be advisable to do this bofore constructing is given below. In drawing up this table, any high energy accelerator of high all fission products with a lifetime of less intensity in order to be sure not to run into than a day were neglected. a «barrier of radiation» which would make 1 the servicing of the machine impossible. ∑ σν ratio for different elements. A ν ν REFERENCES 1 Element A, gr ∑σν, mb σν A ∑ 1. Bruninx E. High Energy Reaction Cross-Sections. ν CERN 61-1 and 62-9, Geněve. H 1 0 0 -27 2. Barbier M.,Dutrannois J. Collection Be 9 15 2.3×10 of data for spectroscopic analysis of induced C 12 10 1.2 » activity, CERN 62-22, Geneve. O 16 10 1 » 3. B 1 a s e r J.-P., Barbier M., Dutrannois Mg 24 42 79 » J. In: Proceedings of the International Al 21 22 32 » Conference on Sector-Focused Cyclotrons and Si 28 38 61 » Meson Factories (CERN, Apr. 1963). Mn 55 51 43 » 4. Barbier M. La radioactivité induite dans les Fe 56 72 25 » matériaux par les protons, neutrons et photons Co 59 160 68 » de haute énergie, Industries Atomiques, Revue Internationale Cu 63 95 36 » pour les Applications Pacifiques de Zn 64 221 57 » l'Energie Nucléaire, Editions René Kister, 33 Ta 181 755 5 » quai Wilson, Geněve, No. 5-6, Vol. VII, 1963.

1392